Using acados to redo 3D quad tracking experiments w ppo

Author: pizarrob

.gitignore

@@ -14,7 +14,7 @@ z_docstring.py
 TODOs.md
 
 **/c_generated_code/**
-**/acados_**
+**/acados_*
 
 
 # macOS users

experiments/mpsc/models/rl_models/quadrotor_3D/track/ppo/mpsf/config.yaml

@@ -53,15 +53,15 @@ output_dir: ./models/rl_models/quadrotor_3D/track/ppo/mpsf/
 overrides:
 - ./config_overrides/quadrotor_3D/ppo_quadrotor_3D.yaml
 - ./config_overrides/quadrotor_3D/quadrotor_3D_track.yaml
-- ./config_overrides/quadrotor_3D/nl_mpsc_quadrotor_3D_linear.yaml
+- ./config_overrides/quadrotor_3D/nl_mpsc_quadrotor_3D.yaml
 restore: null
 safety_filter: nl_mpsc
 seed: 2
 sf_config:
   cost_function: one_step_cost
   decay_factor: 0.85
-  horizon: 5
-  integration_algo: LTI
+  horizon: 20
+  integration_algo: rk4
   mpsc_cost_horizon: 2
   n_samples: 6000
   prior_info:
@@ -85,8 +85,9 @@ sf_config:
   - 90
   slack_cost: 250
   soften_constraints: true
+  use_acados: true
   use_terminal_set: false
-  warmstart: false
+  warmstart: true
 tag: temp
 task: quadrotor
 task_config:

experiments/mpsc/models/rl_models/quadrotor_3D/track/ppo/mpsf/logs/loss/approx_kl.log

@@ -1,101 +1,101 @@
 step,loss/approx_kl
-10000,0.019218411521675687
-20000,0.026657083141617476
-30000,0.01746212083380669
-40000,0.009919317167562742
-50000,0.02111920174987366
-60000,0.01724486625753343
-70000,0.01632642283414801
-80000,0.015934750298038126
-90000,0.0311812112107873
-100000,0.02331677962404986
-110000,0.03371710733820995
-120000,0.030351283960044377
-130000,0.03279125356736282
-140000,0.013648889834682145
-150000,0.023243115593989686
-160000,0.018773438936720295
-170000,0.018916703512271246
-180000,0.020599004921192925
-190000,0.02473791797334949
-200000,0.030151523572082322
-210000,0.0300494180371364
-220000,0.022491315736745794
-230000,0.02257620928188165
-240000,0.0337772676255554
-250000,0.02179785487242043
-260000,0.024567173871522147
-270000,0.023649742904429637
-280000,0.03622848563827573
-290000,0.032509950951983536
-300000,0.025151518592610955
-310000,0.03157009757123887
-320000,0.022510625546177228
-330000,0.02347554156246285
-340000,0.03789273059616486
-350000,0.018296154340108235
-360000,0.024769589103137455
-370000,0.03000542554073036
-380000,0.027254434209316968
-390000,0.019858964625746014
-400000,0.021827833106120433
-410000,0.024644354016830523
-420000,0.024889172101393343
-430000,0.025677736227711044
-440000,0.017630940380816657
-450000,0.017418021506940322
-460000,0.03001865083351732
-470000,0.03849459284295638
-480000,0.027187612513080238
-490000,0.02783633583846191
-500000,0.030406417387227214
-510000,0.02205329760909081
-520000,0.017947127064689997
-530000,0.015077587123960257
-540000,0.028686774040882785
-550000,0.023915739000464478
-560000,0.03114956431090831
-570000,0.03602626387340327
-580000,0.03039917517453432
-590000,0.031206621027862026
-600000,0.022293894877657292
-610000,0.028865009934330978
-620000,0.02208978966809809
-630000,0.028119065674642722
-640000,0.021514978787551323
-650000,0.026450703401739394
-660000,0.03150768955238163
-670000,0.016960548066223657
-680000,0.026525620510801668
-690000,0.03453690588163834
-700000,0.01535425819456577
-710000,0.0443646027551343
-720000,0.028342895085612933
-730000,0.031697115022689105
-740000,0.020270801335573196
-750000,0.02270585975299279
-760000,0.029777690395712847
-770000,0.01838145664272209
-780000,0.018685533913473287
-790000,0.024748957948759195
-800000,0.025136707412699855
-810000,0.020907793069879214
-820000,0.0339710091240704
-830000,0.03614181123363475
-840000,0.037953167936454216
-850000,0.023853258819629748
-860000,0.02006980065877239
-870000,0.03955576866865158
-880000,0.03445271523669362
-890000,0.021048704162240025
-900000,0.004874991532415152
-910000,0.027964027691632504
-920000,0.029088548819224035
-930000,0.01806592782959342
-940000,0.02563529678930839
-950000,0.021795741065094867
-960000,0.029152677084008854
-970000,0.027936897209535037
-980000,0.03480880114560326
-990000,0.02758963455756505
-1000000,0.019105506129562855
+10000,0.019267342627669375
+20000,0.024363837763667108
+30000,0.014531421413024265
+40000,0.015404991634810966
+50000,0.01967445264259974
+60000,0.011831035542612275
+70000,0.015947554850329955
+80000,0.012423110799863935
+90000,0.015280927252024411
+100000,0.02220213850960135
+110000,0.012349804987510047
+120000,0.01828191550448537
+130000,0.028774391937380035
+140000,0.015258365931610266
+150000,0.02345449778561791
+160000,0.027410029154270894
+170000,0.019542859680950642
+180000,0.019825634950151046
+190000,0.03221142959470551
+200000,0.034784734916562844
+210000,0.030562105650703116
+220000,0.03874263330362737
+230000,0.021048107215513784
+240000,0.023490131444608174
+250000,0.021312880609184508
+260000,0.016016186385725936
+270000,0.02354920314004024
+280000,0.02117116949521005
+290000,0.02161381165497005
+300000,0.019996917573735118
+310000,0.025390521561106043
+320000,0.019103998783975836
+330000,0.023890269749487443
+340000,0.02149774165203174
+350000,0.027560550440102815
+360000,0.019251088829090195
+370000,0.018506101751700043
+380000,0.028167877004792292
+390000,0.03479318170187374
+400000,0.0266167871033152
+410000,0.015584384184330701
+420000,0.023963576337943476
+430000,0.02370765237137675
+440000,0.0196848229970783
+450000,0.01707787923514843
+460000,0.022351718166222174
+470000,0.029980707820504902
+480000,0.01593708001698057
+490000,0.0208636624738574
+500000,0.01895671679327885
+510000,0.02578317878457407
+520000,0.023863097761447228
+530000,0.024787811158845822
+540000,0.031156336826582746
+550000,0.022468352271243934
+560000,0.026645291860525812
+570000,0.024720640030379094
+580000,0.01759668216109276
+590000,0.02706180753496786
+600000,0.0339090638483564
+610000,0.02853210787288844
+620000,0.025294201184685033
+630000,0.02775376898547014
+640000,0.025818154526253544
+650000,0.029325115183989205
+660000,0.030363137849296125
+670000,0.033694293132672706
+680000,0.030928333460663752
+690000,0.01576555427163839
+700000,0.017671081035708386
+710000,0.026511379843577744
+720000,0.00761902214338382
+730000,0.022826692027350264
+740000,0.049169556067014744
+750000,0.031370429896439116
+760000,0.02555970596149563
+770000,0.02349021608630816
+780000,0.02183050780246655
+790000,0.02328371241067847
+800000,0.02474643061868846
+810000,0.031218589407702287
+820000,0.031579947207743916
+830000,0.023295751493424176
+840000,0.02685841120158633
+850000,0.023855931901683412
+860000,0.022539945039898156
+870000,0.024766629251341023
+880000,0.020198249692718186
+890000,0.026560715958476068
+900000,0.016740009219696123
+910000,0.021413230057805772
+920000,0.026581356239815557
+930000,0.0316499317996204
+940000,0.02264405793199937
+950000,0.025666885481526458
+960000,0.01787282700339953
+970000,0.02585337059572339
+980000,0.026152768296500047
+990000,0.01878221851463119
+1000000,0.02222854793071747